Great work here, Mike. Tapping into industry leaders is something substack really makes easy, and I applaud you including Conrad's work/ideas.
Let me ask you for your own conclusion: do you think any of these six companies has legit "commercial legs"? I think Runway has been getting the most buzz lately, but I can't really get a sense of whether it will become the dominant paradigm, or even a notable competitor. Any thoughts on this?
Look at how Microsoft talks about BAAI, they are world class.
I think Pi will evolve into something special.
I am very bullish on people like Contextual A.I. and ElevenLabs that deserve to be add-on to this list. Yes any company working with foundational models has a lot of potential since it's still so early.
Cohere and Anthropic are way more legit on the Enterprise side by the end of 2023 than many realize. Just because OpenAI is a first mover doesn't really mean anything, having a huge budget helps but does not guarantee dominance here.
What's missing? The Chinese equivalent of these startups. Don't expect the U.S. alone to dominate. I'm less bullish on Runway, since Generative A.I. has a long ways to go in video and gaming. But in coding is making huge progress.
So there's about three dozen startups that I'm really watching.
Thanks, Mike! I agree completely with your assessment that it's super duper early, and a first mover "advantage" can actually be more of a disadvantage, as more nimble firms move into newly opened spaces and new markets.
The China AI situation is something I've written about; I think there's a real change they have conceded the race (for the foreseeable future) because of their much more limited, restrictive posture... but you cannot sleep on a 1.4 billion user base, something Open AI lusts after today.
This is a news space that is very unlikely to be boring any time soon.
By the way Andrew, what do you make of OpenAI scraping Reddit comments to train their model and Reddit actually wanting to get paid? It's a really curious situation now on Reddit.
Great question. This is a complex situation that people seem to be quickly jumping to one side or the other on, and I can't do that. I love that it drives straight to the heart of AI and copyright, probably the thing I have found myself writing about the most (or maybe 2nd most; the way our language is evolving really ties in here). I am very interested to see how it plays out, and it will no doubt set a much, much bigger standard.
Great work here, Mike. Tapping into industry leaders is something substack really makes easy, and I applaud you including Conrad's work/ideas.
Let me ask you for your own conclusion: do you think any of these six companies has legit "commercial legs"? I think Runway has been getting the most buzz lately, but I can't really get a sense of whether it will become the dominant paradigm, or even a notable competitor. Any thoughts on this?
Look at how Microsoft talks about BAAI, they are world class.
I think Pi will evolve into something special.
I am very bullish on people like Contextual A.I. and ElevenLabs that deserve to be add-on to this list. Yes any company working with foundational models has a lot of potential since it's still so early.
Cohere and Anthropic are way more legit on the Enterprise side by the end of 2023 than many realize. Just because OpenAI is a first mover doesn't really mean anything, having a huge budget helps but does not guarantee dominance here.
What's missing? The Chinese equivalent of these startups. Don't expect the U.S. alone to dominate. I'm less bullish on Runway, since Generative A.I. has a long ways to go in video and gaming. But in coding is making huge progress.
So there's about three dozen startups that I'm really watching.
Thanks, Mike! I agree completely with your assessment that it's super duper early, and a first mover "advantage" can actually be more of a disadvantage, as more nimble firms move into newly opened spaces and new markets.
The China AI situation is something I've written about; I think there's a real change they have conceded the race (for the foreseeable future) because of their much more limited, restrictive posture... but you cannot sleep on a 1.4 billion user base, something Open AI lusts after today.
This is a news space that is very unlikely to be boring any time soon.
By the way Andrew, what do you make of OpenAI scraping Reddit comments to train their model and Reddit actually wanting to get paid? It's a really curious situation now on Reddit.
Great question. This is a complex situation that people seem to be quickly jumping to one side or the other on, and I can't do that. I love that it drives straight to the heart of AI and copyright, probably the thing I have found myself writing about the most (or maybe 2nd most; the way our language is evolving really ties in here). I am very interested to see how it plays out, and it will no doubt set a much, much bigger standard.